DESCRIPTION (provided by candidate): Parent-child interactions often become less predictable during the early adolescent transition period. While there has been much research on the content of these interactions, far less is known about the relative stability or variability of these behavior patterns. The current study will test the hypothesis that the variability in early adolescent interactions is evidence of a phase transition, a dynamic systems concept that refers to periods of fluctuation observed when one stable pattern breaks down and another emerges in its place. This study has four specific aims: (1) to test the hypothesis that flexibility or variability in parent-adolescent interactions peaks during the early adolescent transition phase; (2) to test the hypothesis that high negative emotional intensity decreases flexibility in these dyadic interactions, especially at the transition phase; (3) to examine individual differences in these developmental profiles and test the hypothesis that when puberty and a school transition coincide there are more extreme changes in flexibility than when they occur sequentially; (4) to apply and refine a new dynamic systems method for the study of developmental changes observed in real time. Mothers and daughters will be video-taped in problem-solving discussions at each of three waves: before, during, and after the transition to middle school (grade 7), which approximately coincides with puberty. Grid-like maps of behavior will be created from the coded video-taped observations and several measures of flexibility will be derived and compared across this developmental period.